>many IE-Semitic
> correspondences are treated like valid Nostratic
> comparanda, ....more often than not, they are evidence of
> loans rather than inherited words.
You're probably right. But there are non-loanable points of similarity
which interest me (or would interest me if PIE did not claim any time I have
for this stuff).
Original two-consonant roots, extended to three consonants.
Vowels do not carry significant lexical information; consonants do.
Three types of stop consonants.
Distinction of active and stative declensions.
Original use of a case system, at first probably nom, acc, gen.
Poorly developed system of prefixes.
I'm sure the list could be extended. A glimpse at say, Maori or Chinese,
shows how different these two language groups are from many others. Yet the
evidence seems to suggest any connection between Semitic and PIE is remote -
not that I know the evidence myself well enough to assess it properly.
Peter